flat-pebble conglomerate: its multiple origins and...

Flat-pebble conglomerate: its multiple origins and relationshipto metre-scale depositional cycles

PAUL M. MYROW*, LAUREN TICE*, BONNY ARCHULETA� , BRYN CLARK� ,JOHN F. TAYLOR§ and ROBERT L. RIPPERDAN–

*Department of Geology, The Colorado College, 14 E. Cache La Poudre, Colorado Springs, CO 80903,USA (E-mail: [email protected])�Department of Geology, Washington State University, Pullman, WA 99164, USA�Department of Geology and Geophysics, University of Wyoming, Laramie, WY 82071-3006, USA§Department of Geosciences, Indiana University of Pennsylvania, Indiana, PA 15705, USA–Department of Earth and Atmospheric Sciences, Saint Louis University, St Louis, MO, USA

ABSTRACT

Carbonate flat-pebble conglomerate is an important component of Precambrian

to lower Palaeozoic strata, but its origins remain enigmatic. The Upper

Cambrian to Lower Ordovician strata of the Snowy Range Formation in

northern Wyoming and southern Montana contain abundant flat-pebble

conglomerate beds in shallow-water cyclic and non-cyclic strata. Several

origins of flat-pebble conglomerate are inferred for these strata. In one case, all

stages of development of flat-pebble conglomerate are captured within storm-

dominated shoreface deposits of hummocky cross-stratified (HCS) fine

carbonate grainstone. A variety of synsedimentary deformation structures

records the transition from mildly deformed in situ stratification to buckled

beds of partially disarticulated bedding to fully developed flat-pebble

conglomerate. These features resulted from failure of a shoreface and

subsequent brittle and ductile deformation of compacted to early cemented

deposits. Failure was induced by either storm or seismic waves, and many

beds failed along discrete slide scar surfaces. Centimetre-scale laminae within

thick amalgamated HCS beds were planes of weakness that led to the

development of platy clasts within partly disarticulated and rotated bedding of

the buckled beds. In some cases, buckled masses accelerated downslope until

they exceeded their internal friction, completely disarticulated into clasts and

transformed into a mass flow of individual cm- to dm-scale clasts. This

transition was accompanied by the addition of sand-sized echinoderm-rich

debris from local sources, which slightly lowered friction by reducing clast–

clast interactions. The resulting dominantly horizontal clast orientations

suggest transport by dense, viscous flow dominated by laminar shear. These

flows generally came to rest on the lower shoreface, although in some cases

they continued a limited distance beyond fairweather wave base and were

interbedded with shale and grainstone beds. The clasts in these beds show no

evidence of extensive reworking (i.e. not well rounded) or condensation (i.e.

no rinds or coatings). A second type of flat-pebble conglomerate bed occurs at

the top of upward-coarsening, metre-scale cycles. The flat-pebble conglomerate

beds cap these shoaling cycles and represent either lowstand deposits or, in

some cases, may represent transgressive lags. The clasts are well rounded,

display borings and have iron-rich coatings. The matrix to these beds locally

includes glauconite. These beds were considerably reworked and represent

condensed deposits. Thrombolites occur above the flat-pebble beds and record

microbial growth before initial transgression at the cycle boundaries. A third

Sedimentology (2004) 51, 973–996 doi: 10.1111/j.1365-3091.2004.00657.x

� 2004 International Association of Sedimentologists 973

type of flat-pebble conglomerate bed occurs within unusual metre-scale, shale-

dominated, asymmetric, subaqueous cycles in Shoshone Canyon, Wyoming.

Flat-pebble beds in these cycles consist solely of clasts of carbonate nodules

identical to those that are in situ within underlying shale beds. These deeper

water cycles can be interpreted as either upward-shoaling or -deepening

cycles. The flat-pebble conglomerate beds record winnowing and reworking of

shale and carbonate nodules to lags, during either lowstand or the first stages

of transgression.

Keywords Cambrian, depositional cycles, flat-pebble conglomerate, Montana,Snowy Range Formation, Wyoming.

INTRODUCTION

Carbonate intraclastic rudstone to floatstone(Embry & Klovan, 1971), herein referred tocollectively as flat-pebble conglomerate, is abun-dant within Precambrian to lower Palaeozoiccarbonate-rich successions. There are severalpublished accounts of modern examples ofvertically imbricated flat pebbles (Dionne, 1971;Sanderson & Donovan, 1974) and shells (Green-smith & Tucker, 1968, 1969; Sanderson & Dono-van, 1974). Although some ancient flat-pebbleconglomerate is well sorted and shows verticalpacking of clasts (e.g. Roehl, 1967), most is poorlysorted and tends to have relatively flat-lying clastorientations compared with modern examples(McKee, 1945; Jansa & Fischbuch, 1974;Kazmierczak & Goldring, 1978; Chow, 1986;Myrow et al., 2003). Numerous authors havehypothesized about its origin and the reason(s)for its stratigraphic restriction to older strata (e.g.Sepkoski, 1982). Although petrographic andmacroscopic textural features of flat-pebble con-glomerate have been used to construct plausiblemechanisms for its origin (e.g. combined flows:Mount & Kidder, 1993), the highly variabledepositional processes that created these bedshave remained elusive. This is due, in part, to thefact that transitional development of these bedshas not been documented from ancient deposits.It also results from the paucity of flat-pebbleconglomerate in modern deposits, and thus a lackof uniformitarian models. This reflects differentgeotechnical and diagenetic characteristics ofPrecambrian to early Palaeozoic sediment, com-pared with that of the rest of the Phanerozoic.Such conditions, which included the absence ofabundant and deep-burrowing organisms thatchurned marine sediment after the Ordovicianradiation (Sepkoski, 1982; Droser & Bottjer, 1989),imparted unusual geotechnical properties to thesediment at that time (Bottjer et al., 2000). This

study examines the sedimentological and strati-graphic aspects of flat-pebble conglomerate bedsin the uppermost Cambrian (upper SunwaptanStage and lower Skullrockian Stage) in the nor-thern Rocky Mountains and demonstrates fourorigins of these deposits and their depositionalmechanics.

The Middle Cambrian to Lower Ordovicianstratigraphy of Montana and Wyoming has beenstudied for more than 100 years and given aplethora of lithostratigraphic names (Peale, 1890,1893; Blackwelder, 1918; Deiss, 1936, 1938; Dorf& Lochman, 1940; Mills, 1956; Goodwin, 1961;Martin et al., 1980). Two of the more frequentlyapplied formation names for the uppermostCambrian and basal Ordovician strata in thissuccession are the Gallatin Formation and SnowyRange Formation. The latter term is used in thispaper because several of the key sections arewithin or near the type areas of that formationand its members as they were thoroughly charac-terized by Grant (1965). These strata were thesubject of pioneering biostratigraphic work, re-gional mapping studies and examined as part ofexploration efforts for oil and gas, but few modernprocess-oriented sedimentological studies havebeen undertaken on these units. The basal MiddleCambrian transgressive Flathead Sandstone isoverlain by a succession of shale- and lime-stone-rich units that lie below a prominent Mid-dle Ordovician unconformity at the base of theBighorn Formation. The Upper Sunwaptan toSkullrockian Snowy Range Formation provides arelatively complete succession of trilobite andconodont zones that span the Cambrian–Ordovi-cian boundary interval (Grant, 1965). It containslithofacies typical of the inner detrital belt ofwestern North America (Lochman-Balk, 1971),namely shale, grainstone and flat-pebble con-glomerate. Time-equivalent strata in central andwestern Colorado contain similar lithologies andrepresent a southern extension of the same inner

974 P. M. Myrow et al.

� 2004 International Association of Sedimentologists, Sedimentology, 51, 973–996

detrital belt ( Myrow et al., 2003). The shorelineof this belt is not easily constrained, given recentdetailed stratigraphic studies indicating thatmany tectonic elements of the TranscontinentalArch may not have developed until the EarlyOrdovician (Runkel, 1994; Myrow et al., 2003).Thus, shale-rich deposits of the inner detrital beltmay have covered much of the Rocky Mountainregion and other areas of central Laurentia duringdeposition of the Snowy Range Formation in thelatest Cambrian.

Exposures of the Snowy Range Formationexamined in this study are located across nor-thern Wyoming and south-central Montana in theBighorn, Absaroka, Big Snowy and Pryor Moun-tains (Fig. 1). Lithofacies of the Snowy RangeFormation are described and interpreted below toprovide a stratigraphic and sedimentological

framework for the analysis of depositionalmechanics of these flat-pebble conglomerate beds.

FACIES DESCRIPTIONS

Shale facies

This facies consists of dark grey to greenish-greyshale. Bed thicknesses range from millimetrescale to 60 cm and average 7 cm thick. The shaleis commonly interbedded with grainstone andflat-pebble conglomerate beds (Fig. 2A and B). AtShoshone Canyon (Fig. 1), this facies containsrounded carbonate nodules that range in shapefrom irregular spheres to flattened discs in shale-dominated cycles.

Grainstone facies

This facies is composed of grey to reddish-greyweathering grainstone. Fine-grained, well-sortedgrainstone is most abundant, although beds ofcoarse-grained, bioclast-rich grainstone are alsocommon. Beds range from £1 cm to 130 cm thick,but average � 4 cm. Petrographic study revealsthat the fine grainstone is well laminated, withconsiderable variability in composition betweenlaminae. It consists of fine sand-sized bioclasticgrains and 10–30% peloids. Siliciclastic-richlaminae contain up to 45% silt-sized quartz andfeldspar grains (average 20–25%) and 15% glauc-onite grains, whereas carbonate-rich laminaeaverage � 5% quartz/feldspar and 1–2% glaucon-ite. The quartz and feldspar silt grains arevariably rounded, but most are subangular toangular. Coarser grainstone (coarse to very coarsegrains) generally contains abundant brachiopodand echinoderm fragments and less quartz/feld-spar (generally < 5%) and glauconite (< 2%).Both fine and coarse grainstones generally con-tain fine-grained syntaxial sparry cement.

Grainstone beds are commonly interlayeredwith shale beds of subequal thickness (Fig. 2Aand B). Most grainstone beds are tabular andlaterally extensive, although some pinch-and-swell and others are discontinuous. Grainstonelocally forms isolated gutter casts up to severalcentimetres thick within shale beds. Very thingrainstone lenses also form isolated starved sym-metrical ripples within shale beds (Fig. 2A).Sedimentary structures include wave- and cur-rent-generated, ripple-scale cross-stratification,hummocky cross-stratification (HCS) and parallellamination (Fig. 2B).

Fig. 1. Location map of field sites in Wyoming andMontana. BSM, Careless Creek, Big Snowy Mountains;CB, Cookstove Basin; EPM, East Pryor Mountain; MM,Medicine Mountain; PCE, East Post Creek; PCW, WestPost Creek; SPW, Sheridan Pass; SCW, Shoshone Can-yon; TSC, Tensleep Canyon.

Flat-pebble conglomerate 975


Amalgamated hummocky cross-stratified andparallel-laminated grainstone beds form units upto 1Æ5 m thick. There are two types of HCS: (1) atypical style with 1 cm breaks that define tabularlaminae (Fig. 2C); and (2) an unusual style withuneven, pinch-and-swell lamination (Fig. 2D).

Flat-pebble conglomerate facies

This facies consists of 1–75 cm thick beds ofgrey, green and brown weathering flat-pebbleconglomerate with variable amounts of grain-stone matrix. This facies has a diverse set ofcharacteristics that depend in part on the strat-igraphic context. Fabrics range from clast tomatrix supported (Fig. 3A and B) and gradecontinuously into grainstone beds with scatteredflat pebbles. Bed geometries range from tabular topinch-and-swell to lenticular, and bed contactsvary from relatively planar to highly irregular.Flat-pebble clasts are tabular, and most arecomposed of very fine-grained grainstone

identical in composition and texture to that ofthe fine-grained grainstone facies. Matrix-richbeds generally contain much smaller, mostlypebble-sized clasts only 2–20 mm long. Clast-rich beds contain larger, flat, ellipsoidal pebblesto cobbles 30–150 mm long (Fig. 3C). Most flat-pebble clasts are 5–20 mm thick in cross-section(average � 10 mm). Clasts at the tops of somebeds contain abundant borings (Fig. 3E). Thedegree of rounding is highly variable from highlyrounded to angular clasts that in some casesshow partial brittle deformation.

Petrographically, most clasts are identical incharacter to the grainstone facies. They are lam-inated and generally contain between 10% and35% quartz and feldspar silt and up to 5%glauconite grains. The matrix of the conglomeratevaries in grain size and composition, but it iscoarser than that found within clasts on average(Fig. 3C), has more abundant brachiopod andechinoderm fragments and contains less quartz/feldspar (5–10% average). The glauconite and

Fig. 2. (A) Very thin-bedded grainstone and shale. Note abundant starved symmetrical ripples. (B) Thin-beddedgrainstone and shale. Grainstone beds are highly irregular in thickness and show wave-ripple lamination, small-scaleHCS and parallel lamination. (C) Amalgamated HCS grainstone. Note low-angle truncation surfaces and cm-scalebreaks in lamination. (D) Unusual pinch-and-swell HCS lamination within amalgamated grainstone. Pencil is 14 cmlong.



peloid content is comparable to that found withinclasts. Flat pebbles at the top of some beds areseparated by white, thrombolitic boundstone(Fig. 3E).

Intraclasts are variably oriented, but they aresubhorizontal in most beds, particularly in bedswith higher grainstone matrix content (Fig. 3B).

Isolated vertically imbricated rosettes (smallfans of clasts) exist both within and on top ofbeds (Fig. 3D), but these are extremely rare.Many beds contain evidence of amalgamation,such as changes in average clast size, degree ofclast support, sorting and grainstone content.Some flat-pebble beds contain large thrombolite

Fig. 3. (A) Clast-supported flat-pebble conglomerate. Note dominantly horizontal orientations of flat pebbles.(B) Close-up of grainstone-supported flat-pebble conglomerate. Matrix consists of coarse echinoderm-rich grainstone.(C) Bedding plane view of flat-pebble conglomerate bed with large rounded flat pebbles. Note large star-shapedechinoderm ossicles (e). (D) Flat-pebble conglomerate bed with dominantly horizontal clast orientations. A largerosette of imbricated clasts extends above the general level of the bed. (E) Top of flat-pebble conglomerate bed withthrombolite matrix (t) between the clasts. Note the circular borings on many of the clasts. (F) Large (55 · 20 cm) raftof laminated grainstone within a flat-pebble conglomerate bed. Pencil is 14 cm long.



clasts up to 24 cm across, and a few flat-pebblebeds contain large grainstone blocks up to60 · 20 cm across (Fig. 3F). In a few cases,flat-pebble conglomerate beds grade laterallyinto brittly and ductilely deformed, thicklylaminated carbonate grainstone, referred tobelow as buckled beds.

Thrombolite facies

This facies consists of discrete masses of white,grey and yellow weathering thrombolite bound-stone (Fig. 4). Thrombolites range up to 44 ·95 cm in cross-section. Many weather massively,but some exhibit a characteristic mottled internal

Fig. 4. (A) Thrombolite (to right of pencil) and adjacent flat-pebble bed with local thrombolite matrix.(B)Thrombolite mound (t) with abundant flat pebbles surrounded by flat-pebble conglomerate (FP). This layer isunderlain by very thin grainstone and shale beds and overlain by shale. (C) White-weathering thrombolite mound(43 cm thick) among grainstone and minor shale beds. (D) Close-up of thrombolite mound with typical mottledfabric. (E) Channel of grainstone (g) cut into bioturbated carbonate mudstone with overhanging thrombolite mound(t) along the edge of the channel (arrowed). (F) Thick (78 cm) channel fill of cross-bedded grainstone with white-weathering thrombolite along the edge of the ancient channel. Pencil is 14 cm long.



fabric (Fig. 4D). Thrombolites form mounds on topof, and locally within, flat-pebble conglomeratebeds (Fig. 4A and B). Thrombolite mounds locallygrade upwards from underlying flat-pebble con-glomerate beds through a zone of flat pebbles withthrombolitic carbonate matrix. Some thrombolitescontain flat pebbles that range from a few scatteredclasts to highly abundant (Fig. 4B). Thrombolitesrest on the tops of tabular beds but also overliechannel surfaces (Fig. 4E and F and 5), particularlyalong their steep edges, where they exhibit over-hanging geometries (Fig. 4E).

Buckled bed facies

This facies ranges from mildly disturbed, platy,fine-grained grainstone beds to clast-supportedflat-pebble conglomerate with relict, partly disar-ticulated bedding (Fig. 6). Beds of this facies rangefrom 5 cm to 80 cm thick and average about 30 cmthick. Clasts range from 1 to 20 cm in length. Theleast deformed beds have gentle irregular foldsand minor brittle deformation. More intenselydeformed beds show abundant breakage of lam-ination into flat-pebble clasts that exist in domainswith fitted fabrics. These domains record rotationand both ductile and brittle deformation ofcm-scale grainstone laminae. Although the fabricsmay locally resemble rosettes produced by wavereworking of loose clasts, the fabric is much morefitted and simply results from local vertical rota-tion and fracturing of bedding. The platy clastsand deformed, relatively intact laminae havesharp edges reflecting brittle deformation,although some of both are gently folded (Fig. 6).Isolated clasts have more rounded edges. Theresulting fabrics are highly irregular and changelaterally over short distances. Buckled beds areintimately interlayered and interfingered with

thick HCS grainstone beds (Figs 6A, D and F and7). These transitions range from subtle and grada-tional to sharp along truncation surfaces. Thesesurfaces truncate as much as 55 cm of HCSgrainstone (Fig. 7). The buckled beds are boundedby highly irregular contacts and thus abruptlythicken and thin laterally due to a combination ofmounding at their tops and erosion along theirbases (Fig. 7). Hummocky cross-stratified grain-stone beds locally onlap buckled beds (Figs 7 and6F). Thrombolitic boundstone fills the spacesbetween the clasts at the top of a few buckledbeds at East Pryor Mountain, just as it occurs at thetop of some non-deformed flat-pebble beds. Inseveral cases, the truncated upturned edges ofbuckled bedding are encrusted by echinodermholdfasts up to 1Æ8 cm in diameter (Fig. 6E). Inseveral localities, buckled beds are laterally trans-itional into flat-pebble conglomerate beds.Although these transitions are in many cases toopoorly exposed to examine thoroughly, severalexamples are documented below.

Facies associations

Two facies associations are recognized in theSnowy Range Formation. A mixed facies associ-ation consists of all the facies described above,namely shale, starved ripples to thin beds of finegrainstone, coarse grainstone, flat-pebble con-glomerate beds, buckled beds and thrombolites.Shale content varies but generally ranges from30% to 60% of these deposits. Most of the SnowyRange (� 90%) consists of this mixed faciesassociation. The second, grainstone-rich faciesassociation consists of thick intervals of HCS finegrainstone, buckled beds and flat-pebble con-glomerate beds. Shale is notably absent or is lessthan 5% of this facies association.

Fig. 5. Channel-fill deposits at 26Æ0 m in measured section at Cookstove Basin, WY. Note thrombolite mounds alongedge of channel. Thrombolites also grew in part on flat-pebble conglomerate fill.



PALAEOENVIRONMENTALINTERPRETATION

The Snowy Range Formation was deposited in anepicratonic setting, within the inner detrital beltof Laurentia (Lochman-Balk, 1971; Sepkoski,

1982). Hence, maximum water depth was prob-ably several tens of metres. The grainstone-richfacies association with thick units of amalgama-ted HCS grainstone and near absence of shalereflects deposition in a shoreface setting, i.e.above fairweather wave base (cf. Walker, 1984).

Fig. 6. (A) Buckled bed with irregular base overlying amalgamated HCS grainstone. (B) Buckled bed that is almosttransitional to flat-pebble conglomerate bed. Bedding is nearly intact to right of pencil. (C) Bedding plane view ofbuckled bed with echinoderm holdfasts on upturned and broken grainstone bedding. (D) Buckled bed that overlies aflat-pebble bed and underlies more shaly facies. (E) Transition from parallel-laminated grainstone (g) to buckledgrainstone (b). Note highly gradational transition. Hammer for scale. (F) Buckled grainstone within thick unit ofamalgamated HCS grainstone (see Fig. 7). Note irregular top and base of buckled unit. Arrows point to onlap ofgrainstone above the buckled mass. Field of view ¼ 1Æ4 m. Pencil in other photos is 14 cm long.



Fig.7.

Ph

oto

mosa

ican

dcorr

esp

on

din

gfi

eld

sketc

hof

am

alg

am

ate

dH

CS

gra

inst

on

ean

dass

ocia

ted

bu

ckle

dbed

sat

Post

Cre

ek

East

locali

ty(1

1–12

m).

Note

dow

ncu

ttin

gat

base

of

bu

ckle

dbed

din

g(s

lid

esu

rface)

an

don

lap

of

gra

inst

on

eabove.



Abundant fine-grained HCS grainstone indicatesthat the shoreface existed in a storm-dominatedsetting, and parallel lamination indicates thatthese storm flows reached upper flow regimeplane bed conditions. High-resolution strati-graphic correlation of numerous outcrops inWyoming and Montana is under way, and itappears that shoreface environments may havebeen patchily distributed in space and time. Inother words, large sand bodies might have beenscattered throughout the inner detrital belt and,whether or not they were completely emergent,the processes of deposition and geomorphologicalcharacter of these sand bodies indicate depositionabove fairweather wave base in a manner identi-cal to that of a shoreface of a single linearshoreline.

Abundant shale beds in the mixed faciesassociation indicate deposition in a low-energysetting below fairweather wave base. A near totalabsence of desiccation cracks in these depositsplaces this facies association in a more seawardpalaeoenvironment than that of the shorefacedeposits of the grainstone-rich association. Well-sorted, fine-grained grainstone beds were depos-ited predominantly by storm-generated flows, asindicated by the presence of gutter casts, sym-metrical ripples and HCS, features common totempestites (Kreisa, 1981; Myrow, 1992). Well-sorted, fine-grained carbonate sand was presum-ably eroded from the shoreface and transportedoffshore during storms. Interbedded shale andstarved symmetrical ripples of fine grainstonerepresent the most distal deposits, whereasthicker fine grainstone beds with well-developedHCS interbedded with less shale represent moreproximal deposits (i.e. standard tempestite proxi-mality trends: Walker, 1984; Aigner, 1985).

The difference in grain size between the quartzand feldspar grains (silt) and the carbonate grains(fine sand on average) in the grainstone of theSnowy Range Formation probably reflects depos-ition of the siliciclastic component by aeolianprocesses. The compositional (� 20% feldspar)and textural immaturity of the siliciclastic com-ponent of the grainstone may reflect one or morefactors: (1) relatively minor abrasion in aeoliantransport because of low mass (silt-sized grains);(2) arid climatic conditions; and (3) an origin asfirst-generation grains that have undergone littletransport in fluvial environments. Given theconsiderable thickness of Cambrian strata thatunderlie the Snowy Range across the northernRocky Mountain region, the siliciclastic grainswere either derived from local unrecognized

highlands or were carried long distances by windand deposited across the inner detrital belt.Angular quartz silt is described from the shalyhalf-cycles of Grand Cycles by Chow & James(1987) and, in their interpretation, the strata weredeposited on the flanks of carbonate shoal com-plexes that were located on the seaward side of anintrashelf basin far from cratonic sources ofquartz silt. Aeolian transport of siliciclastic siltmay have been important for many such UpperCambrian intrashelf basins.

The thrombolite facies was deposited in clearshallow water as a result of the buildup of smallmicrobial mounds. In some sections, the throm-bolitic facies occurs as the highest and shallowestfacies in shale-based, metre-scale subtidal cycles.These cycles and their significance with regard tochanges in relative sea level are described later inthis paper. The origin of buckled beds and flat-pebble conglomerate beds is the main focus ofthis paper and is discussed in detail below.

Synsedimentary deformation structures andrelationships to flat-pebble conglomerate

A variety of mass flow and synsedimentarydeformation features exist within the SnowyRange Formation, including slumps, slide scarsand thrust-duplicated beds. Of particular import-ance are transitions from buckled beds to flat-pebble conglomerate, which are exposed at anumber of sections. At Post Creek East (Fig. 1),amalgamated HCS grainstone grades laterally intoa buckled bed that in turn grades laterally into aflat-pebble conglomerate bed with echinoderm-rich grainstone matrix and subhorizontally orien-ted flat-pebble clasts (Fig. 8). The gradationaltransition from the buckled bed to the flat-pebbleconglomerate is subtle. A second transitionbetween buckled grainstone and flat-pebble con-glomerate occurs less than 4 m higher in the samesection (Fig. 9). Here, two large rafts of thin-bedded grainstone occur within this interval. One86 · 25 cm raft rests on buckled bedding but isotherwise surrounded by flat-pebble conglomer-ate (Fig. 9). The other raft (60 · 20 cm across) ispartly covered at one end but appears to beencased within flat-pebble conglomerate.

At Medicine Mountain (Fig. 1), a prominentslide scar truncates amalgamated grainstone, and

Fig. 8. Field sketch from Post Creek East locality(12Æ61 m) that highlights features of buckled beds, flat-pebble conglomerate beds and the transition betweenthe two facies.



the sides of the scar are overlain by buckledgrainstone. The buckled deposits grade laterallyinto flat-pebble conglomerate. At East PryorMountain (Fig. 1), a complex buckled and flat-

pebble conglomerate unit has an upper zone ofbuckled beds covering a slide scar that cutsamalgamated grainstone. Below the slide scar,the grainstone grades into buckled grainstone and

Fig. 9. Photomosaic and corresponding field sketch of transition from buckled bedding to flat-pebble conglomerateat Post Creek East locality (16Æ36 m). Note large raft of parallel-laminated fine grainstone resting on buckled beddingand surrounded by flat-pebble conglomerate.



then flat-pebble conglomerate. Transitionsbetween amalgamated grainstone and buckledgrainstone are generally gradational and showall degrees of deformation of amalgamated grain-stone from minor disruption to extensive brittleand ductile deformation. Transitions frombuckled bedding range from gradational to sharp.Even where the transition is covered, buckledgrainstone and flat-pebble conglomerate bedsoccur laterally along the same stratigraphic hori-zon. In some cases, the ends of the buckled bedsare abruptly truncated, and adjacent beds consistof shale and grainstone that were depositedsubsequent to failure and removal of buckledgrainstone.

In all the cases described above, the flat-pebbleconglomerate lacks much of the evidence forphysical reworking and condensation describedearlier (e.g. extensive rounding, clast coatings,borings), which is common to beds that capshoaling cycles.

INTERPRETATION

The common continuum of amalgamated HCSgrainstone into buckled bedding within the grain-stone-rich facies association indicates that failureof the shoreface gave rise to brittle and ductiledeformation of these compacted to early cemen-ted deposits. Failures were both gradational, asevidenced by lateral transitions with buckledbeds, or more catastrophic and localized, asevidenced by slide scars that mark sharp faciesboundaries. These failures were probably also inpart gravity-driven, although on a regional scaleslopes were almost negligible in the epicratonicsetting of the inner detrital belt ( Myrow et al.,2003). Equilibrium slopes on the ancient carbon-ate sand shorefaces were probably steeper (1� ormore) as a result of sediment buildup by strongwave and storm influence.

The buckled beds appear to be surface phe-nomena, based on a number of criteria. First,buckled beds form mounded structures withsnouts that are similar to those of debris flows.Snouts are onlapped by overlying beds and thusdo not appear to have been intrastratally insertedinto surrounding bedding. Secondly, echinodermholdfasts are attached to the broken edges ofclasts at the top of numerous buckled beds.Thirdly, thrombolitic microbial communities col-onized the tops of the buckled and flat-pebblebeds and also occupied open space betweenclasts, the latter indicating growth of microbes

within open-framework crypts at the depositionalinterface. Fourthly, within thick shoreface depos-its of the grainstone-dominated facies association,buckled beds that rest on slide scars are onlappedby draping grainstone (Fig. 7), indicating that thebuckled beds were surface features with relief oftens of centimetres.

Sediment removed along slide scars was depos-ited laterally as buckled beds or slumps and, insome cases, became involved in transitions to flat-pebble conglomerate. A model for these transi-tions is shown in Fig. 10. Buckling presumablyresulted from external body forces followed bydownslope movement along the shoreface. Itinvolved brittle deformation of early cementedlayers and folding of less cemented, but well-compacted bedding. The cm-scale laminationdeveloped within the amalgamated HCS providednatural breaks along planes of weakness thatdirectly yielded cm-thick, platy, early cementedclasts. These clasts are preserved as partiallydisarticulated and rotated bedding within thebuckled beds. Friction between thin, large clastskept the bedding relatively intact. Once down-slope acceleration of slump masses exceeded theinternal friction of the slump masses, they brokeinto disarticulated clasts, resulting in a mass flowof individual cm- to dm-scale clasts. These clastspresumably moved as a dense, friction-dominateddebris flow to yield flat-pebble conglomerate. Inother cases, clasts that formed by slumping mayhave been extensively reworked by waves andcurrents and then deposited as typical flat-pebblebeds with rounded clasts and evidence of con-densation.

The development of flat-pebble conglomeratefrom failure of thick amalgamated grainstoneunits is supported by the presence of large raftsof bedded grainstone within flat-pebble beds(Fig. 9). These rafts reflect wholesale failure ofthe shoreface and incorporation of large coherentblocks into mass flows, probably as a result offailure along slide scars that are well developedin the amalgamated grainstone facies. This trans-ition to debris flow was in most cases alsoaccompanied by the addition of sand-sized ech-inoderm-rich debris (Fig. 10). This debris isnotably much coarser that the fine grainstone thatmakes up the amalgamated HCS shoreface depos-its. Thus, this debris was carried shoreward byshoaling waves and/or currents from deeperwater settings where stalked echinoderms grew(Fig. 10), presumably in response to the sameforcing mechanism (e.g. storm) that caused failureof the shoreface. Mixing of coarse carbonate sand



with flat-pebble debris probably lowered overallfriction by allowing tabular clasts to slip alongsmall rounded grains and thus reduced clast–clast interactions. The dominantly horizontalfabric of these flat-pebble beds (e.g. Fig. 8) alsosupports the interpretation of these as mass flowdeposits. This fabric is less likely in subaqueoussettings influenced by storm-generated waves andcurrents, where incremental buildup of clastswould probably have formed strong edgewisefabrics including wave-generated rosettes (Dion-ne, 1971; Sanderson & Donovan, 1974; Mount &Kidder, 1993). The dominantly horizontal fabricis considered to be a result of viscous flowdominated by laminar shear. These friction-dom-inated flows accelerated down the front of the

shoreface and generally came to rest on the lowershoreface, thus producing the common associ-ation of amalgamated grainstone, buckled bed-ding and flat-pebble conglomerate. However,some flows may have continued a limited dis-tance beyond the shoreface, as indicated by thepresence of some flat-pebble conglomerate bedsinterbedded with shale and grainstone that recorddeposition beyond the lower shoreface transition.Emplacement by such a mechanism explains anenigmatic aspect of these flat-pebble conglomer-ate beds, namely the fact that many of the beds aretabular (on the scale of metres to tens of metres)and rest directly on shale with relatively planarlower surfaces. This is consistent with laminarflow and argues against in situ reworking by

Fig. 10. Model for the formation offlat-pebble conglomerate along awave/storm-dominated carbonateshoreline. Stage I: conditions beforestorm. Stage 2: cyclic wave loadingduring storm causes failure of thefine grainstone and downslopemovement, resulting in buckled bed.This is accompanied by shorewardtransport of locally derived echino-derm debris. Stage 3: disarticulationof clasts in buckled bed and mixingwith transported echinoderm debrisleads to mass flow and deposition ofa flat-pebble conglomerate withsubhorizontally oriented clasts in acoarse bioclastic matrix. Stage 4:post-storm colonization of hardsurfaces, including buckled bed andflat-pebble bed.



waves and currents that would be expected toproduce highly irregular depressions on a muddysurface. Deposition of buckled and flat-pebblebeds produced abundant hard surfaces for colon-ization by stalked echinoderms (Fig. 10).

The mechanism(s) for the proposed failures ofthe shoreface are difficult to discern. Disruptionof strata by body forces is generally linked tofailure induced by shaking created by earth-quakes or the passage of storm-generated waves.Distinguishing between these causal mechanismsis extremely difficult and, in most ancientdeposits, it is not possible. Purported seismitesgenerally consist of in situ liquefaction featuressuch as sedimentary dikes, convolute beddingand ball-and-pillow (Pope et al., 1997; Oberme-ier, 1998; Obermeier & Pond, 1999), and there isa surprising lack of such features in the rocks ofthis study.

Storm waves have long been considered astriggers for failure of marine sediment (e.g.Henkel, 1970; Prior & Coleman, 1982; Clukeyet al., 1985; Prior et al., 1989). Failure resultswhen fluctuations in bottom pressure associatedwith the passage of large surface waves createcyclic shear stresses, which in turn cause anincrease in pore fluid pressure and associateddecrease in strength, leading to failure anddownslope movement. Such failure has beensuggested for the mobilization of fine sand fromshoreface settings during storms (Walker, 1984)and for soft-sediment deformation features result-ing from liquefaction, as mentioned above (Moli-na et al., 1998). However, the shorefacegrainstone deposits of this study show evidencefor early compaction and cementation in the formof brittle deformation structures, so it is unlikelythat these deposits could have had enough porefluid to fail by liquefaction. The transmission ofpressure waves generated by storms could none-theless have caused failure of these deposits andthe initial movement of buckled masses. Thetransport and mixing of coarse echinoderm grainsfrom offshore with flat pebbles during theseevents is more compatible with storm-generatedfailure and transport than with the passage ofseismic waves.

The flat-pebble conglomerate deposits des-cribed above are an extraordinary example inwhich transitional stages of development werecaptured in the rock record, and these allow forthe reconstruction of depositional processes. Flat-pebble beds also occur in three other types ofcyclic and non-cyclic deposits within the SnowyRange Formation, and these are described below.

These demonstrate the multiplicity of origin offlat-pebble conglomerate within even a singleformation, and emphasize the importance of thestratigraphic and sedimentological context of flat-pebble facies for their proper interpretation.

FLAT-PEBBLE CONGLOMERATE BEDSAS SHORELINE DEPOSITS

Description

The youngest (highest Sunwaptan and basal Skull-rockian) strata of the Snowy Range Formation inmost localities constitute a relatively resistant,carbonate-rich unit that forms a cliff directly belowthe base of the Bighorn Formation, The base of thisrelatively resistant unit (Fig. 11; 24Æ05–26Æ65m) isin relatively sharp contact with an underlyingmuch shalier facies that contains several thromb-olite-capped cycles (described below) at CookstoveBasin (Fig. 1). The uppermost part of this cyclicinterval is a 53 cm thick unit of shale with< 1–3 cm thick grainstone beds, including starvedwave ripples and shrinkage cracks in the top10 cm. Directly above is an anomalously thick(75 cm) flat-pebble conglomerate with verticallyimbricated flat pebbles across the top layer and inmultiple microdomains within the bed (Fig. 12b).The flat-pebble conglomerate is in turn overlain by68–90 cm of thinly to thickly laminated grainstonewith parallel lamination and HCS (Fig. 12a). A0–22 cm thick lenticular flat-pebble conglomeratebed with a concave-up basal erosion surface capsthis grainstone. Both this lenticular flat-pebble bedand the underlying/adjacent grainstone bed areoverlain by a discontinuous 0–14 cm thick bed ofcoarse grainstone and iron-coated flat pebbles withborings. This bed has an irregular upper surfacethat is in sharp contact with 76 cm of shale withvery thin to thin beds of grainstone (� 50% shale).

A second, 1 m thick, amalgamated unit offlat-pebble conglomerate exists at 30Æ8 m at Cook-stove Basin (Fig. 11). This unit also rests on ametre-scale shoaling cycle with thin-beddedgrainstone and mud-cracked shale directly belowthe flat-pebble unit. The flat-pebble conglomeratedisplays a variety of fabrics from flat-lying claststo vertically oriented rosettes.

Interpretation

The strata directly above 24Æ05 m have manyfeatures diagnostic of a storm-influenced shore-line. The amalgamated, parallel-laminated and



HCS grainstone beds are diagnostic of a wave-and storm-influenced shoreface environment, asdiscussed earlier. The stratigraphic transition at25Æ65 m to distal shaly facies with starved ripplesrepresents a fairly profound flooding surface. Thethin, lenticular coarse grainstone and flat-pebblebeds that mark this transition represent a lagdeposited on a ravinement surface caused byshoreface retreat associated with transgression.The iron-coated flat pebbles with borings in thelag deposit are analogous to the hardgrounds

described by Osleger & Read (1991). The chemicalreactions that create these coatings and theabundance of borings probably resulted from arapid decrease in the rate of sediment accumula-tion across the flooding surface.

The stratigraphic position of the thick flat-pebble conglomerate (24Æ05 m) below the

Fig. 12. (A) Thick amalgamated flat-pebble conglo-merate bed and overlying amalgamated HCS bed atCookstove Basin, WY. Sledgehammer for scale. (B)Close-up of transition between the flat-pebble con-glomerate and HCS beds. Note thin layer of verticallyimbricated pebbles at the top of the flat-pebble bed.Lens cap is 5Æ5 cm across.

Fig. 11. Part of stratigraphic section from CookstoveBasin, WY. A transition from beach flat-pebble con-glomerate to shoreface grainstone to shale-rich offshoredeposits is captured between 24 and 26 m. A secondthick flat-pebble conglomerate unit of beach origin oc-curs between 31 and 32 m. Scale in metres. T-shapedsymbols to left of column indicate desiccation cracks.Sh, shale; Fn Gr, fine grainstone; Cr Gr, coarse grain-stone; Fpc, flat-pebble conglomerate.



shoreface deposits of the amalgamated grainstonefacies suggests that it represents an environmentnear and above mean low tide. It formed in asimilar shoreline position to the underlyingshale-rich deposits, which were subaerially ex-posed and probably represents a mud-flat envi-ronment. This change records a shift towardsstrong wave and storm activity, as evidenced bythe rosettes of edgewise clasts. In these examples,the flat-pebble conglomerate represents either abeach setting or less continuous flat-pebble shoalson a mud flat. The transition from the flat-pebbleconglomerate to the amalgamated grainstone rep-resents a marine flooding surface. It is unclearwhether the flat-pebble foreshore deposits formedas a transgressive lag or from reworking duringlowstand conditions just before relative sea-levelrise.

The second unit of flat-pebble conglomerate at30Æ8 m is similarly interpreted as a beach deposit.It also rests on shale with desiccation cracks andconsists of multiple beds of flat pebbles. The flat-pebble units at the base of the relatively resistantuppermost Snowy Range (Fig. 11) are anomal-ously thick relative to others within the forma-tion. In addition, both 24Æ05 and 30Æ8 m beds reston mud-cracked intervals, which are extremelyrare within these inner detrital belt deposits.

SHOALING CYCLES

Flat-pebble beds in the Snowy Range Formationoccur most commonly interbedded with shaleand grainstone beds without any apparent verti-cal stratigraphic trends in facies or grain size.This is somewhat remarkable given the generallyshallow-water setting of the Cambrian–Ordovi-cian inner shelf lagoon of Laurentia and the factthat nearshore to shoreline deposits of this age arecommonly cyclic (Aitken, 1978; Markello & Read,1982; Osleger & Read, 1991, 1993). However,upward-coarsening cycles (Fig. 13) do occur inthe upper part of the Snowy Range Formation atsome localities. The cycles vary in degrees ofcompleteness, consist of three or four facies andrange in thickness from 0Æ95 to 1Æ89 m. Shoalingcycles are best developed at Careless Creek(Fig. 1) where 10 cycles occur within < 10 m ofsection (Fig. 14). Each cycle begins with 4–23 cmof shale that in some cases contains interbeddedmillimetre-scale grainstone laminae. The shale tograinstone ratio is � 9:1. Above these units are 6–51 cm of intercalated 1–8 cm thick grainstonebeds and very thin shale beds. The shale

percentage decreases upwards to � 10–15%. Thegrainstone beds tend to pinch-and-swell, andsome have wave-rippled upper surfaces(Fig. 13C). The tops of the cycles consist of flat-pebble conglomerate beds ranging from 5 to26 cm thick (Fig. 13A and B). The beds arelaterally continuous and tabular, although some

Fig. 13. (A) Shoaling cycle with thick lower unit ofshale (Sh) with very thin grainstone beds, a thingrainstone bed (g) and a capping flat-pebble conglom-erate. (B) Shoaling cycle with shale (Sh), grainstone (g)and flat-pebble conglomerate (FP) beds. Cycleoverlies flat-pebble bed at top of underlying cycle.(C) Thrombolite bed (T) overlain by the base of shoal-ing cycle. The cycle illustrates the upward decrease inshale and increase in grainstone content and bedthickness. Pencil is 14 cm long.



have mildly erosional bases. The clasts in thesebeds are highly rounded, locally contain abun-dant borings and, in some cases, display dark

brown iron-rich coatings, as seen in beddingplane exposures. The matrix of these beds is richin coarse bioclasts, particularly echinoderm frag-

Fig. 14. Measured section with upward-shoaling cycles. The Careless Creek section contains 10 upward-shoalingcycles (arrowed), none of which contains thrombolite caps. The Cookstove Basin section shows four cycles, three ofwhich are capped by prominent thrombolites. Scale in metres. Sh, shale; Gr, grainstone; Fp, flat-pebble conglomerate;Th, thrombolite. Idealized cycle is shown on bottom.



ments, and in many cases contains glauconitegrains.

Thrombolites are absent from the cycles atCareless Creek (Fig. 1) but cap flat-pebble con-glomerate beds in similar cycles elsewhere(Fig. 13C). Many of these thrombolitic moundsgrade upwards from the flat-pebble conglomeratethrough a zone of clast-supported flat pebbleswith thrombolitic boundstone filling much of thespace between them. Variations from the ideal-ized cycle include the absence of thromboliticmounds, some interbedding of grainstone withflat-pebble conglomerate at the base of the flat-pebble beds and the presence of buckled beds inplace of (or laterally adjacent to) the flat-pebblebeds. Additionally, the lower shale units in somecases contain thicker grainstone beds up to 1 cmthick (Fig. 13C). The top of one cycle at TensleepCanyon (Fig. 1) is marked by a 36 cm thick bedof very coarse crinoidal grainstone with abun-dant flat pebbles instead of a flat-pebble con-glomerate bed. The upper 10 cm of this coarsegrainstone contains thrombolite clasts and over-lying thrombolite mounds up to 8 cm thick thatcap the cycle.

Interpretation

The upward-coarsening cycles of the SnowyRange Formation contain no evidence for subaer-ial exposure such as shrinkage cracks, palaeokarstor other features indicative of meteoric vadose orphreatic diagenesis, and are thus inferred to bewholly subaqueous. The upward increase in bedthickness and percentage of fine grainstonereflects shoaling and a shift to shallower waterconditions. The grainstone beds may in somecases represent deposition in a lower shoreface.The flat-pebble conglomerate beds that cap thesecycles may represent foreshore deposits or near-shore deposits (below fairweather wave base)transported from the shoreline either duringlowstand by storms or during subsequent trans-gression by various marine processes. The tabularnature of many of the beds and their relatively flatlower surfaces (Fig. 13) argue against in situerosion of underlying grainstone as a source ofthe clasts, but instead suggest some transport anddeposition of the flat pebbles. Evidence for suchtransport includes the rounded nature of theclasts. These beds also contain evidence of con-densation in the form of borings, clast coatingsand glauconite grains. The general lack of chaoticfabrics and edgewise rosettes (cf. Mount & Kidder,1993) brings into question the depositional

mechanics of these beds. Edgewise fabrics aregenerally attributed to waves and combined-flowstorm currents, but little or no experimental workhas been done to determine whether dominantlyhorizontal fabrics could also form in bedloadtransport of flat pebbles under such conditions.These questions are also important for the inter-pretation of flat-pebble beds that rest on shale innon-cyclic strata in the Snowy Range Formation,as these beds also lack edgewise fabrics and are inmany cases tabular.

Numerous authors have described peritidalmetre-scale carbonate cycles, and these com-monly have flat-pebble conglomerate andthrombolites at their bases and overlying cryp-talgal laminite with evidence of desiccation (seereview by James, 1984). These are generallyinterpreted as having subtidal bases and inter-tidal tops. The flat-pebble conglomerate beds atthe base are commonly interpreted as transgres-sive lags (e.g. James, 1984; Wright, 1984; Grot-zinger, 1986; Tucker & Wright, 1990). The cyclesdescribed herein are shale based and entirelysubaqueous. They resemble, in scale and faciesarrangement, the shale-based cycles described byMarkello & Read (1982) and Osleger & Read(1991). The latter interpreted their cycles ashaving aggraded from below to above storm wavebase, with the flat-pebble conglomerate bedsrepresenting reworking of underlying grainstone.A similar interpretation is likely for the SnowyRange cycles, but it is suggested here that thetransition from shale with grainstone to amalga-mated grainstone near the top of these cycles mayinstead represent a transition across fairweatherwave base and the lower shoreface transitionzone. The palaeoenvironmental setting of thethrombolites is also problematic. In the cyclesdescribed by Demicco (1983) and Koerschner &Read (1989), microbial mounds are considered tobe subtidal to lower intertidal features. In thisstudy, there is no evidence for subaerial expo-sure, and thus the thrombolites are considered tobe entirely subaqueous, although they probablyformed during the late stage of shoaling and thusrepresent the shallowest facies (cf. Markello &Read, 1982). The thrombolite growth was sub-strate controlled by a shift to the hard sea floorsurfaces underlain by flat pebbles. Thrombolitefacies described by Saltzman (1999) from slightlyolder rocks in the field area also rest directlyon flat-pebble conglomerate and attest to thissubstrate control. No independent evidence(high-resolution geochronological data) has beencollected to demonstrate that these cycles were



driven by eustasy, but an origin from high-frequency sea-level fluctuations is likely.

SHALE-DOMINATED CYCLES

A second, more enigmatic type of asymmetriccycle is found in the Shoshone Canyon section(Figs 1, 15 and 16), where five cycles exist within5 m of section. The cycles commence with 10–32 cm of shale with relatively abundant carbonatenodules and, in some cases, thin grainstone beds.

These nodule-rich shale units directly overlie flat-pebble conglomerate beds and are overlain by 26–60 cm units of shale with sparse nodules. Pebbleconglomerate beds containing clasts of carbonatenodules overlie these shale units. The pebbleconglomerate beds range from 8 to 34 cm thick,are generally laterally continuous and havemildly irregular upper and/or lower surfaces.

Fig. 16. Shale-rich cycles from Shoshone Canyon, WY.(A) Two shoaling cycles. Lower cycle has a muchthinner intraclast conglomerate bed at the base. (B)Close-up of upper cycle in (A). Cycle consists of basalintraclast bed (I), shale with carbonate nodules (N) andshale without nodules (N). Upper intraclast bed is baseof another cycle. Hammer for scale in (A) and (B). (C)Close-up of intraclast bed. Note that intraclasts arerounded carbonate nodules similar to those interbed-ded with the shale within the cycles. Pencil for scale.

Fig. 15. Part of a stratigraphic section from ShoshoneCanyon, WY, showing five grainstone-poor cycles(arrowed). Scale in metres. Sh, shale; Gr, grainstone;Fp, flat pebble conglomerate.



Interpretation

The asymmetric shale-dominated cycles fromShoshone Canyon can be interpreted as eithershoaling cycles or deepening cycles (Fig. 17). Ineither case, they are considered to be whollysubaqueous, as they are shale rich and contain noevidence for subaerial exposure. These cyclesdiffer from the upward-coarsening cycles des-cribed earlier in two ways. First, carbonate nod-ules are present within the shale facies and arethe sole source of clasts within the pebbleconglomerate. Secondly, these cycles containvery little grainstone and no thrombolites.

If the cycles record shoaling (Fig. 17), the topsurface of the flat-pebble conglomerate bedswould represent the bases of the cycles. Rapidflooding of a flat-pebble bed and deposition ofmud would have initiated these cycles. Carbon-ate nodules were precipitated and, in somecases, sparse thin lime sand beds were depos-ited. If one assumes that precipitation of carbon-ate nodules is a function of time in which thesediment is in a shallow, geochemically distinctlayer in the sediment column, then nodulegrowth would be in part a function of theaccumulation rate of mud. Lower rates wouldprolong the time that a particular layer ofsediment would be in the shallow diageneticzone in which nodules grow, whereas higherrates would tend to bury a layer below the zoneof nodule precipitation. If the basal nodule-richunits at the bases of these cycles record relative

sea-level rise, they may reflect alluviation and ageneral decrease in sediment delivery to theshelf. In this case, nodule precipitation wasenhanced by reduction in mud accumulationrates. The upward change from shale withabundant nodules to shale with sparse noduleswould represent a fall in relative sea level,increased sediment accumulation rates and low-er rates of nodule growth. Continued relative sea-level fall would have resulted in deposition ofnearshore/shoreline pebble deposits. The factthat there are no grainstone beds at the top ofthese cycles, and that the intraclastic conglom-eratic beds are not composed of flat pebblesderived from thin grainstone beds, but insteadreworked nodules, indicates that the shorelinewas too distal or not sandy during deposition ofthese cycles. Reworking by nearshore processesin these grainstone-starved settings would haveinvolved removal of mud and production of anodule lag. The excavation and winnowing ofearly cemented carbonate nodules is similar tothat described by Mount & Kidder (1993),although the flat-pebble beds they describe restdirectly on nodule-rich strata. In the presentstudy, shale directly underlying the intraclastconglomerate beds contains relatively few nodu-les, so it is likely that laterally, more nodule-richlayers were being reworked and transportedsome distance to the site of deposition. Thepresence of thin grainstone beds within lowershale beds in three of the cycles is problematicin this hypothesis.

Fig. 17. Alternative hypotheses for grainstone-poor cycles.



A second hypothesis is that these representupward-deepening subaqueous cycles (Fig. 17).In this case, the cycles would begin with near-shore processes reworking shale with carbonatenodules to form intraclastic conglomerate. Theflat-pebble beds could represent either a lowstanddeposit or a transgressive lag. Relative sea-levelrise would cause mud to be deposited on top ofthe intraclastic conglomerate. In this case, thingrainstone beds within shale in three of the cycleswould represent a transition from limited near-shore sources of grainstone to deeper settings thatare more distal from a source of grainstone. Theshale with very few nodules might represent amaximum flooding interval. The reason for thestratigraphic distribution of carbonate nodules isnot readily obvious in this hypothesis. Shallow-marine, metre-scale deepening cycles are notcommon features in the rock record, and so theshoaling hypothesis is considered to be the morelikely of the two. In either case, the general lack ofgrainstone in these shale-dominated cycles mayreflect deposition in deeper water environmentsaway from the shoreline sources of grainstone,although the production of intraclast conglomer-ate requires that shoaling took place to a level atwhich marine processes could rework depositsinto a nodule lag.

SUMMARY

Flat-pebble conglomerate is an important compo-nent of carbonate deposits, particularly in Cam-brian and Lower Ordovician strata (Sepkoski,1982). Flat-pebble accumulations have been des-cribed from modern supratidal and beach envi-ronments (e.g. Roehl, 1967; Dionne, 1971;Sanderson & Donovan, 1974), but most authorshave concluded that the vast majority of ancientflat-pebble conglomerate beds were deposited insubtidal environments below fairweather wavebase (Kazmierczak & Goldring, 1978; Markello &Read, 1982; Sepkoski, 1982; Demicco, 1983;Wilson, 1985; Grotzinger, 1986; Lee & Kim,1992; Mount & Kidder, 1993). Given that subaer-ial diagenetic processes were not responsible forlithification of the carbonate beds, it was conclu-ded early on that submarine cementation orhardground formation was a necessary step inthe formation of flat pebbles. Sepkoski (1982)argued that the evolution of metazoa at the base ofthe Cambrian resulted in abundant faecal pelletsthat were a presorted source of thin grainstonebeds. The thin pelletal limestone beds were

highly permeable and easily cemented so thatthe reworking of these early cemented beds wasthe source of flat pebbles. There has been muchspeculation concerning the processes by whichsuch beds were reworked into flat-pebble con-glomerate, but few detailed process-oriented sed-imentological studies have been undertaken onflat-pebble facies.

Flat-pebble conglomerate beds of the SnowyRange Formation of northern Wyoming andsouthern Montana are shown here to be ofvariable origin and to occur stratigraphicallywithin both non-cyclic and cyclic strata. Fourtypes of flat-pebble beds are described. The firsttype described here is shown to have resultedfrom failure and mass movement of thicklylaminated strata of carbonate sand shorefacedeposits of the grainstone-dominated faciesassociation. Such deformation resulted fromeither storm-generated or seismic waves. Buck-ling, rotation and disarticulation of cm-scalelamination within HCS grainstone led to thedevelopment of viscous flows dominated bylaminar shear. The addition of coarse grainstonereduced friction in these flows and may haveallowed them to move beyond the shorefaceacross the muddy bottom of the proximal near-shore zone. Many hypotheses have been offeredover the years for the origin of flat-pebble con-glomerate, including deposition from mass flows,but this study is the first to demonstrate aparticular origin using deposits that show allstages of development of such beds.

The origins of the other three types of flat-pebblebeds were derived from analysis of fully devel-oped flat-pebble conglomerate beds. An anomal-ously thick amalgamated flat-pebble bed at onesection (Cookstove Basin) can be interpretedstraightforwardly as a beach deposit that is over-lain by a thick shoreface deposit of amalgamatedHCS grainstone, then a flooding surface and anoverlying shale-rich unit. The interpretation of theflat-pebble bed, with an upper surface of wave-generated pebble rosettes, within this transgres-sive succession is relatively straightforward.

A third type of flat-pebble conglomerate bedsoccurs at the tops of metre-scale, subtidal toperitidal, shoaling cycles. Sedimentological ana-lysis of the cycles indicates that the flat-pebblebeds formed by wave and/or storm reworking ofshoreline deposits during lowstand conditions oras a transgressive lag during the earliest stages ofsubsequent flooding at the base of the next cycle.A fourth type of pebble conglomerate, consistingof reworked carbonate nodules, occurs within



enigmatic, shale-dominated, metre-scale cycles.Although it is unclear whether these representupward-deepening or upward-shoaling cycles, theconglomerate beds are lags formed by the winnow-ing and reworking of shale with abundant carbon-ate nodules by wave and/or storm processes.

This study demonstrates the multiple origins offlat-pebble conglomerate beds within a singleunit, the Upper Cambrian to Lower Ordovicianstrata of the Snowy Range Formation. The inter-pretation of flat-pebble conglomerate beds there-fore requires careful analysis of their stratigraphicand sedimentological context. These must beunderstood within the history of secular changesin palaeogeography, biological evolution, oceanicchemistry and the geotechnical properties ofsediment.

REFERENCES

Aigner, T. (1985) Storm Depositional Systems. Lecture Notes

in Earth Sciences 3. Springer-Verlag, New York, 174 pp.

Aitken, J.D. (1978) Revised models for depositional Grand

Cycles, Cambrian of the southern Rocky Mountains,

Canada. Can. Petrol. Geol. Bull., 25, 515–542.

Blackwelder, E. (1918) New Geological formations in western

Wyoming. Washington Acad. Sci. J., 8, 417–426.

Bottjer, D.J., Hagadorn, J.W. and Dornbos, S.Q. (2000) The

Cambrian substrate revolution. GSA Today, 10, 1–7.

Chow, N. (1986) Sedimentology and diagenesis of middle and

upper Cambrian platform carbonates and siliciclastics, Portau Port Peninsula, Western Newfoundland. Unpubl. PhD

Thesis. Memorial University, St John’s, Newfoundland,

458 pp.

Chow, N. and James, N.P. (1987) Cambrian Grand Cycles: a

northern Appalachian perspective. Geol. Soc. Am. Bull., 98,418–429.

Clukey, E.C., Kulhawy, F.H., Liu, P.L.-F. and Tate, G.B. (1985)

The impact of wave loads and pore-water pressure genera-

tion on initiation of sediment transport. Geo-Mar. Lett., 5,177–183.

Deiss, C. (1936) Revision of type Cambrian formations and

sections of Montana and Yellowstone National Park. Geol.

Soc. Am. Bull., 47, 1257–1342.

Deiss, C.H. (1938) Cambrian formations and sections in part of

Cordilleran Trough. Geol. Soc. Am. Bull., 49, 1067–1168.

Demicco, R.V. (1983) Wavy and lenticular-bedded carbonate

ribbon rocks of the Upper Cambrian Conococheague Lime-

stone, Central Appalachians. J. Sed. Petrol., 53, 1121–1132.

Dionne, J.C. (1971) Vertical packing of flat stones. Can. J. Earth

Sci., 8, 1585–1591.

Dorf, E. and Lochman, C. (1940) Upper Cambrian formations

in Southern Montana. Geol. Soc. Am. Bull., 51, 541–556.

Droser, M.L. and Bottjer, D.J. (1989) Ordovician increase in

extent and depth of bioturbation: implications for under-

standing early Paleozoic ecospace utilization. Geology, 17,850–852.

Embry, A.F. and Klovan, J.E. (1971) A Late Devonian (sic) reef

tract on Northeastern Banks Island, N.W.T. Can. Petrol.

Geol. Bull., 19, 730–781.

Goodwin, P.W. (1961) Fauna and stratigraphy of the Cam-

brian and Lower Ordovician of the Bighorn Mountains,

Wyoming. Unpubl. MSc. Thesis, University of Iowa, 73 pp.

Grant, R.E. (1965) Faunas and stratigraphy of the Snowy

Range Formation (Upper Cambrian) in southwestern Mon-

tana and northwestern Wyoming. Geol. Soc. Am. Mem., 96,171 pp.

Greensmith, J.T. and Tucker, E.V. (1968) Imbricate structures

in Essex offshore shell banks. Nature, 220, 1115–1116.

Greensmith, J.T. and Tucker, E.V. (1969) The origin of Holo-

cene shell deposits in the chenier plain facies of Essex

(Great Britain). Mar. Geol., 7, 403–425.

Grotzinger, J.P. (1986) Evolution of Early Proterozoic passive-

margin carbonate platform, Rocknest Formation, Wopmay

Orogen, Northwest Territories. Can. J. Sed. Petrol., 56, 831–

847.

Henkel, D.J. (1970) The role of waves in causing submarine

landslides. Geotechnique, 20, 75–80.

James, N.P. (1984) Shallowing-upward sequences in carbon-

ates. In: Facies Models (Ed. R.G. Walker), 2nd edn, pp. 213–

228, Geological Association of Canada, Toronto.

Jansa, L.F. and Fischbuch, N.R. (1974) Evolution of a Middle

and Upper Devonian sequence from a clastic coastal plain–

deltaic complex into overlying carbonate reef complexes

and banks, Sturgeon–Mitsue area, Alberta. Geol. Surv. Can.

Bull., 234, 205 pp.

Kazmierczak, J. and Goldring, R. (1978) Subtidal flat-pebble

conglomerate from the Upper Devonian of Poland: a

multiprovenant high-energy product. Geol. Mag., 115,359–366.

Koerschner, W.F. and Read, J.F. (1989) Field and modeling

studies of Cambrian carbonate cycles, Virginia Appalachi-

ans. J. Sed. Petrol., 59, 654–687.

Kreisa, R.D. (1981) Storm-generated sedimentary structures in

subtidal marine facies with examples from Middle and

Upper Ordovician of southwestern Virginia. J. Sed. Petrol.,

51, 823–848.

Lee, Y.I. and Kim, J.C. (1992) Storm-influenced siliciclastic

and carbonate ramp deposits, the Lower Ordovician

Dumugol Formation, South Korea. Sedimentology, 39, 951–

969.

Lochman-Balk, C. (1971) The Cambrian of the craton of the

United States. In: Cambrian of the New World (Ed.

C.H. Holland), pp. 79–168. John Wiley and Sons, New York.

McKee, E.D. (1945) Cambrian History in the Grand CanyonRegion. Carnegie Inst. Washington Publ., 1, 230 pp.

Markello, J.R. and Read, J.F. (1982) Upper Cambrian intra-

shelf basin, Nolichucky Formation, southwest Virginia

Appalachians. AAPG Bull., 66, 860–878.

Martin, W.D., Fischer, H.J., Keogh, R.J. and Moore, K. (1980)

The petrology of the limestones of the Upper Gros Ventre

and Gallatin Limestone Formations (Middle and Upper

Cambrian) Northwestern Wind River Basin, Wyoming.

Wyoming Geol. Assoc. Guidebook, 31, 45–48.

Mills, N.K. (1956) Subsurface stratigraphy of the Pre-Niobrara

Formations in the Big Horn Basin, Wyoming: Wyoming

stratigraphy, Part 1. Wyoming Geol. Assoc., pp. 9–22.

Molina, J.M., Alfaro, P., Moretti, M. and Soria, J.M. (1998)

Soft-sediment deformation structures induced by cyclic

stress of storm waves in tempestites (Miocene, Guadalquivir

Basin, Spain). Terra Nova, 10, 145–150.

Mount, J.F. and Kidder, D. (1993) Combined flow origin of

edgewise intraclast conglomerates. Sellick Hill Formation

(Lower Cambrian), South Australia. Sedimentology, 40,315–329.



Myrow, P.M. (1992) Bypass-zone tempestite facies model and

proximality trends for an ancient muddy shoreline and

shelf. J. Sed. Petrol., 62, 99–115.

Myrow, P.M., Taylor, J.F., Miller, J.F., Ethington, R.L.,Ripperdan, R.L. and Allen, J. (2003) Fallen Arches: dis-

pelling myths concerning Cambrian and Ordovician paleo-

geography of the Rocky Mountain Region. Geol. Soc. Am.

Bull., 115, 695–713.

Obermeier, S.F. (1998) Liquefaction evidence for strong

earthquakes of Holocene and latest Pleistocene ages in the

states of Indiana and Illinois, USA. Eng. Geol., 50, 227–254.

Obermeier, S.F. and Pond, E.C. (1999) Issues in using lique-

faction features for paleoseismic analysis. Seismol. Res.

Lett., 70, 34–58.

Osleger, D. and Read, J.F. (1991) Relation of eustasy to

stacking patterns of meter-scale carbonate cycles, Late

Cambrian, USA. J. Sed. Petrol., 61, 1225–1252.

Osleger, D. and Read, J.F. (1993) Comparative analysis of

methods used to define eustatic variations in outcrop: Late

Cambrian interbasinal sequence development. Am. J. Sci.,

293, 157–216.

Peale, A.C. (1890) Tenth Annual Report of the United StatesGeological Survey of the Interior for 1888–1889. Part 1, pp.

130–132.

Peale, A.C. (1893) Paleozoic section in the vicinity of Three

Forks, Montana. US Geol. Surv. Bull., 110, 1–56.

Pope, M.C., Read, J.F., Bambach, R. and Hofmann, J.J. (1997)

Late Middle to Late Ordovician seismites of Kentucky,

southwest Ohio and Virginia: sedimentary recorders of

earthquakes in the Appalachian basin. Geol. Soc. Am. Bull.,109, 489–503.

Prior, D.B. and Coleman, J.M. (1982) Active slides and flows

in underconsolidated marine sediments on the slopes of the

Mississippi delta. In: Marine Slides and Other Mass Move-

ments (Eds S. Saxov and J.K. Nieuwenhuis), pp. 21–49.

Plenum Press, New York.

Prior, D.B., Suhayda, J.N., Lu, N.-Z., Bornhold, B.D., Keller,G.H., Wiseman, W.J., Jr, Wright, L.D. and Yang, Z.-S. (1989)

Storm wave reactivation of a submarine landslide. Nature,

341, 47–50.

Roehl, P.O. (1967) Stony Mountain (Ordovician) and Interlake

(Silurian) facies analogs of recent low-energy marine and

subaerial carbonates, Bahamas. AAPG Bull., 51, 1979–2023.

Runkel, A.C. (1994) Deposition of the uppermost Cambrian

(Croixan) Jordan Sandstone and the nature of the Cambrian–

Ordovician boundary in the Upper Mississippi Valley. Geol.

Soc. Am. Bull., 106, 492–506.

Saltzman, M.R. (1999) Upper Cambrian carbonate platform

evolution, Elvinia and Taenicephalus Zones (Ptero-

cephaliid–Ptychaspid biomere boundary), northwestern

Wyoming. J. Sed. Res., 69, 926–938.

Sanderson, D. and Donovan, R.N. (1974) The vertical packing

of shells and stones on some recent beaches. J. Sed. Petrol.,

44, 680–688.

Sepkoski, J.J., Jr (1982) Flat-pebble conglomerates, storm

deposits, and the Cambrian bottom fauna. In: Cyclic Event

and Stratification (Eds G. Einsele and A. Seilacher),

pp. 371–388. Springer-Verlag, Berlin.

Tucker, M.E. and Wright, V.P. (1990) Carbonate Sedimentol-

ogy. Blackwell, Oxford, 482 pp.

Walker, R.G. (1984) Shelf and Shallow Marine Sands. In:

Facies Models (Ed. R.G. Walker), 2nd edn, pp. 141–170.

Geological Association of Canada, Toronto.

Wilson, M.D. (1985) Origin of Upper Cambrian flat pebble

conglomerates in the Northern Powder River Basin,

Wyoming. 7th SEPM Core Workshop. Golden, CO, pp. 1–50.

Wright, V.P. (1984) Peritidal carbonate facies models: a

review. Geol. J., 19, 309–325.

Manuscript received 28 February 2003; revisionaccepted 14 April 2004.



flat-pebble conglomerate: its multiple origins and...

Documents